A Novel Approach For English-Hindi Cross lingual Summarization

Cross-lingual summarization (CLS) is the process of producing a gist in a specific language for a source text in another language. There is a scarcity of CLS datasets for English to Hindi language summarization. This work aims to introduce a hybrid approach for CLS using abstractive summarization and machine translation. The source text along with the target summary is fed into deep learning model, which produces an abstractive summary, which consists of newly synthesized sentences. Two models one for generating brief summaries and other for detailed summaries were trained. The generated summary will be translated into Hindi language using a pre-trained transformer. These models were evaluated for their summarization and machine translation capabilities and was found to perform well, compared to the baseline models.